Assessing Programming in the Age of Generative AI
Design assessments that reveal understanding, not just a finished code file.
Request instructor accessSee how it worksA practical assessment pattern
Ask students to write a function that calculates an average, test it with a normal list and an empty list, then explain one design decision. Evaluate the explanation and tests alongside the working code.
Assessment design checklist
State the learning objective; include at least one meaningful edge case; ask for a short explanation; grade the result and the reasoning separately.
The challenge
AI-assisted coding makes final output alone an incomplete picture of student understanding. Instructors need multiple sources of evidence.
Design for explanation and iteration
Use short tasks, clear objectives, staged reasoning, debugging, and reflection alongside functional tests.
Use process evidence carefully
Code replay may help identify revision patterns and support conversations with students, but it cannot reliably determine whether AI was used.
Request a semester of instructor access to evaluate the platform. Approval is required before access is activated.
Request a free instructor trial